Researchers at Oregon State University are using satellite imaging to measure damage in Iran. The Conflict Ecology Lab works to assess the effect peace and conflict have on land. The lab has previously done work around Gaza, Ukraine and Sudan. Jamon Van Den Hoek is an associate professor of geography and geospatial sciences at OSU and leads the lab. He joins us to share more on what he’s seeing.
Note: The following transcript was transcribed digitally and validated for accuracy, readability and formatting by an OPB volunteer.
Dave Miller: This is Think Out Loud on OPB. I’m Dave Miller. For 10 years now, researchers at Oregon State University have used satellite imagery to assess the effects that war has on land. They work in what’s called conflict ecology. They’ve studied visual data from Gaza, Ukraine and Sudan. And they’ve recently turned their attention to Iran.
Jamon Van Den Hoek is an associate professor of [geography] and geopolitical sciences at OSU. He leads this research lab and he joins us now. It’s great to have you on Think Out Loud.
Jamon Van Den Hoek: Thanks for welcoming me.
Miller: How do you define conflict ecology?
Van Den Hoek: Conflict ecology really is just a group of people; it’s not a discipline as such. But we work on topics at the intersection of geospatial science, mapping science, satellite imagery, GPS, GIS (geographic information science.) Those are the tools in our toolkit. Traditionally, those have been deployed to things like natural resource management, or in the Pacific Northwest, wildfire monitoring, prevention, for example.
We tend to use those tools exclusively on thinking about armed conflict settings, humanitarian settings, refugee camps, internally displaced people sites. And we use these tools to understand the immediate effects of war – that tends to be things like damage to cities, damage to homes, damage to agriculture and forest – as well as the long-term implications that cascade from that initial conflict. Our research tends to focus on long-term impacts of people being forced out away from their home, outside of their country, into new refugee settings, and thinking about environmental degradation, food security, long-term climate risks in those new sites which sprawl around the world.
Miller: Where do you get your data from?
Van Den Hoek: We almost exclusively use open data, so civilian spacecrafts, imagery that are collected by NASA and European Space Agency, their counterparts. So these data are openly available, publicly accessible repositories of data that are collected, usually once a week. It varies depending on the specific sensor. We use a variety of different sensors in our work. But these images all tell a different side of the story, so we put all these together when needed. And also to be direct, where possible, when the imagery are collected, we also use commercial imagery provided by businesses not in the civilian space necessarily. Those tend to have different capabilities or are collected at different times in different locations. So when possible, we try to kind of bring those in to help support our interpretation. But the main analysis is done on these data that any of your listeners right now could go and download, and start looking at themselves.
Miller: Are you mainly talking about still photos from satellites, or video?
Van Den Hoek: These are all images, not video, that are collected at a given time, at a given location. But because they are taken repeatedly, this is really the value of having satellites in space, that platform, that perspective, it’s not just for the single day. All of that effort to get into space, the value of it is now it’s in orbit, so it’s coming back again and again and again.
We use these images and stitch them together into animations, film strips as such. And in fact, when we do our analysis … which is all automated, we’re not making the judgment call image by image. We have an automated routine that is built out of a lot of interpretation to help us kind of guide our assessment. But we can take that full collection of images that have been collected, going back over a decade, to build out that historical view. So we have this long-term record. And that’s really where our analysis lies, the changes over time that we can see.
So in a sense, we can make a video. It just wouldn’t be a video anyone would probably like to watch. It’s slow changes. There’s not a lot happening until, in our cases, the conflict starts. And then it depends of course where we’re looking, but we start seeing some very pronounced changes that breaks with the past. They’re anomalies, they look like complete unexpected events. And of course, that’s what’s unexpected – it’s urban damage, agricultural damage.
Miller: I want to turn to Iran. How much have you been able to learn about what’s been happening there? And it’s worth saying again, you said sometimes it’s once a week passes that you have access to, but we are now more than a week into this particular war.
Van Den Hoek: I think we’re in the 12th day now. Our analysis, we’re basically spinning up a new approach right now that one of the postdoctoral researchers in the lab, Corey Scher, has been working on. This new approach we’re doing is really a regionwide assessment to detect locations of damage across the region. I should note that this isn’t resolved enough, it’s not so finely detailed. We can’t see individuals in it, we can’t see cars, so there’s arguably less of a concern from the security perspective. This is really broad-scale monitoring looking for large cases of urban damage around the whole region.
The new thing that we’re doing with Iran is really trying to do this in a cohesive way that allows us to stitch together insights over the broad region. So that would include Kuwait, it would include Iraq, other neighboring Gulf states that have also seen damage. And in this case, we’re not trying to confirm what’s already known. We’re trying to lead with this analysis, we’re trying to find new sites that have not been as well documented. And this is quite common in conflict settings. There’s a lot of attention to places where people live, there’s attention to high profile places such as schools or hospitals. That all makes sense, but then the rest of the geography, the rest of the environment often gets left out of these assessments. We’re very keen to broaden the view, so we’re not just looking at infrastructure, we’re looking at much more of a holistic sense of the war impacts across the whole region.
Miller: Have you gained any insights so far as you’re looking region-wide?
Van Den Hoek: Well, our full analysis isn’t quite ready. In the next day or so, I think we’ll be able to start sending this out to journalists and humanitarians whom we’ve worked with before. But the analysis itself, we have looked at some available high resolution imagery that has been publicly released, that journalists have made use of as well, to try to do some very localized assessments. So far, our only deep dive has been on the school damage in Minab in Southern Iran, to try to guide some interpretation about what may have happened there, knowing as little as we do, to be honest.
These are still, of course, revealed things that are coming out in the international media even today. What was the target? What munitions were used? What was the consequence? Who was behind sending the strike in the first place? These are things we as researchers don’t know. What we can see is interpret the aftermath and interpret some of the environmental damages that result. So we’ve supported some interpretations along those lines.
Miller: Are there ever cases where the public imagery that you rely on is delayed or withheld from you?
Van Den Hoek: Well, it’s sort of a yes and no kind of answer. We have the civilian space agency imagery from NASA and ESA. Those are reliably collected and we’re not aware of any intentional withholding that’s occurred there.
What has happened in the past few days, really going back the last five days, is one of the main providers of commercial satellite imagery around the globe, this company called Planet Labs – who’ve just done a fantastic job at providing imagery in other conflict settings, often in cases where they’re the only reliable, consistent provider of imagery – there have been some additional concerns about the security implications of their release of data. [The data] traditionally had just been released to journalists, humanitarians and anyone else that we know were sort of trusted users, so to speak.
Last week, they put a four-day delay on imagery that would be collected across the Gulf states. Iran was not a part of that restriction at the time. But as of yesterday – so this is just the four-day turnabout – now there’s a two-week delay on those imagery being made available. And it’s not just the states surrounding Iran, it’s also Iran. So it’s sort of a Gulf state-wide blackout, if you will, and it’s a two-week delay. We don’t know the consequences of that yet, but we can hazard all of the implications that that will have on humanitarian awareness, the ability of international media to take stock of the damage, of the impact of the war throughout the whole region. Humanitarians trying to provide relief on the ground won’t have access to that either. It’s really an unprecedented move to impose this kind of delay.
Miller: Who would benefit from that kind of delay?
Van Den Hoek: I think the concern comes from the actors who would seek to reuse those imagery to target U.S. interests, allied interests. This is the actual terminology that’s used in the Planet press release. So that’s the thought, that the blackout is coming to prevent that kind of use. Those that benefit from it, we can only really interpret that from thinking about who still will have access to imagery, when the rest of the world that traditionally has relied on Planet Labs for these data, now that that pipeline is cut off. Who benefits is people that still have their own access to imagery. That’s states, that’s people who have their own, we could colloquially call these spy satellites, that can still monitor things or can still defer to other systems.
It’s a really important caveat here, too, that this restriction that Planet Labs is putting in, as well meaning as it is, does not bar the U.S. government from also accessing their imagery. This is a public blackout. So the government can still make use. And there are contracts, of course, between Planet Labs and the government, so the government can still have access to this.
Miller: It’s media, journalists, nonprofits or humanitarian groups who don’t have access?
Van Den Hoek: That’s right. That’s sort of the two-state split.
Miller: I want to turn to the bigger picture here in terms of imagery, whether it’s still photos or videos, of the kind you rely on and the kind that we in the public see. AI-generated imagery is getting more sophisticated and more prevalent every day. How do you know what to trust?
Van Den Hoek: Well, that’s a sliding scale. We work with such raw imagery downlinked from satellite, directly served up on secure servers. We don’t make use of any of this imagery that could be manipulated or fabricated. The challenge, of course, is that the broader public that’s looking at imagery, there could be older imagery that’s reused. This could be cell phone imagery. This could be satellite imagery. It could be new imagery that’s manipulated to look in a certain way. This is something that’s increasingly being documented. It seems like this is quite a new state of AI generation of fake images that are flooding social media. In some cases, news media have picked these up as well. So it’s extremely problematic.
It doesn’t directly get into our workflow so much, but it certainly can taint the broader public awareness and just the direction, the line of thinking, the line of questioning about how we go about monitoring these impacts, how we go about driving attention to one kind of concern versus the other. It’s confusing things quite a bit.
Miller: Where do your students end up working? You’re giving them this grounding in this way of analyzing visual data about the world. Where might they take those tools?
Van Den Hoek: A lot of the graduate students that have finished through the lab, some will go on to graduate to other graduate programs or postdoctoral positions, staying in that academic research. Several have gone on to working with environmental conservation groups, mainly domestically based. The current direction that we’re at now is this real-time monitoring. We’ve only really been doing this for the past two or three years. It’s sort of a new stage of working in this much greater, as you’re sort of hinting at, spillover and intersect with media, open source intelligence, military issues. So I think time will tell where current students are going.
But this still, despite its sort of operational public facing side, [is] very much still a research project. Very much still developing new tools, asking questions that are relevant for academic research, trying to make as good use as we can of the tools and data that are out there designed for other purposes, but we’re repurposing them and using them in this new setting. And that brings with it a lot of prototyping, a lot of uncertainty about whether things are going to play out. But it also leads to a really experimental and, I think, increasingly impactful position, to try to provide perspectives that are not commonly seen, especially coming out from academia.
Miller: Jamon Van Den Hoek, thanks so much.
Van Den Hoek: Appreciate it. Thank you.
Miller: Jamon Van Den Hoek is an associate professor of geography and the director of conflict ecology at Oregon State University.
“Think Out Loud®” broadcasts live at noon every day and rebroadcasts at 8 p.m.
If you’d like to comment on any of the topics in this show or suggest a topic of your own, please get in touch with us on Facebook, send an email to thinkoutloud@opb.org, or you can leave a voicemail for us at 503-293-1983.
